UK battery revenues: seasonality and trading | Mark Meyrick

Right, forgive me if some of the stuff that you're going to see now, you've seen already, I didn't obviously get visibility of other people's presentations, so I'll try and skirt around those slides a bit. So I was asked to talk about how much seasonality affects battery revenues, and you'll find out that battery can do quite a lot of things. And we heard just now about how batteries save the day, and that's because they can do some great stuff, mostly, well the best thing they do is deliver power quickly, and that's what saved the frequency. But they can do trading, they're often in the balancing mechanism, they can provide reserve services, and reserve services in case you're not sure quite how they work, they're a commitment that you give day ahead to turn up for the balancing mechanism. So that is the NISO, so that is the NISO, so that is the NISO, securing a bit more sort of headroom on its balancing actions. And it can do some of the stuff that was mentioned by Lewis earlier, Blackstart, reactive power, and inertia services. And those things are really for the DSOs, and they're not generally procured under some sort of auction, but they tend to be procured a little bit bilaterally. And then the capacity market, which is long-term contracts to just be available at key times. You can't deliver the same megawatt twice, so most of them aren't stackable, but some services you can split into. But what's important about batteries is you'll often see charts like this, in which an analyst from the market, and this happens to be a MODO slide, sorry, Montel. But you can see that you'll often see a slide like this where batteries seem to be at some time of the year earning more money doing one thing than another, and other times the other flip. And so, for example, in this slide, it looked like the balancing mechanism wasn't so great for batteries in late summer, yet earlier in the summer is making a lot of money. But what you've got to really remember is these are very one-dimensional charts, because, of course, what a battery is doing at any point in time is either charging or discharging. And it can do both of those under any of these services. So sometimes when a balancing mechanism looks like it wasn't very attractive, it's because batteries are using the balancing mechanism to charge up. And other times when a wholesale didn't look like it was very attractive, it's because of exactly the same thing, you're charging up in the wholesale market. So what's going to be interesting is working out how people decide which market to charge up in. Similarly, frequency response services are often used for charging and discharging. They go both ways. Oops, I just want to go back, actually, because we are going to be talking on the panel a little bit about top-bottom spreads. So all those numbers across the summer months there were the two-hour top-bottom spreads for what a battery could earn. And as you can see, they were a bit ho-hum for most of the summer. And suddenly, when you get into the shoulder month, when a bit of volatility comes to the market, you're able to start making some quite big returns in the top-bottom spread. And that's the cheapest two hours against the most expensive two hours. And indeed, in the middle of this dunkelflourter at the moment, we're seeing some quite extreme power prices. Those prices are now up to £170, £180, and probably getting in two cycles as well. Maybe a cycle of £180, maybe a cycle of £80. Whereas most of the summer, you're only getting about one cycle out of the battery. So seasonality is really important from that point of view. And it's important also to remember with the battery that revenue is not value. So you shouldn't obsessively run the battery every day when there are going to be better times in the year when you'll get more value out of the cost of cycling. Because cycling does have a cost. So that's a typical two-cycle day from last month. And it is just as power prices are beginning to evolve into two proper peaks. Interestingly, today, the price for tomorrow is just one great big hump through the day. It's like a 10-hour peak. So that's a one-cycle day. Now, this is a little question for everyone in the audience. This is a chart showing the BM offer prices, the BM bid prices, and the M7 intraday price. Now, who would think it's better for the battery, and that's over the month of October, who would think it's better for the battery to be using the balancing mechanism rather than the intraday market? I've got one, two hands, maybe. So might I assume that everybody else thinks the intraday market is better to trade in? Yeah, okay. There's a lot of people sitting on the fence. We're going to... Look, come on, give me a break. I'm trying to wake you up before the drinks. Okay, so we're going to go into this a little bit, because it's not as simple as it looks, as I'm sure you appreciate. And why are... One of the things we wanted to talk about in this session was why revenues do flip-flop around. I can't quite tell you why there is a knife and fork in the skipping circle. There's a good reason for it, and it wasn't down to me, let me tell you. But all the things we've been talking about earlier this afternoon, heat, gas, constraints, a lot of solar, these have all affected why the battery optimiser goes about his business the way they do. So let's have a look at gas. Gas, as was mentioned, I think, by Ben, has been setting the marginal price for 80% of this year. Now, that sounds like a lot, but it was 90% in 2022, 2023, so it is gradually moving away. But, you know what, when gas goes up to those sort of levels, it is going to pull the price up. And, logically, the wider range of prices you have on a day, the greater opportunities for spread trading, cheapest to lowest, except for a day like tomorrow, when, as I said, it's just one long peak. And it's not just in this country, as we've seen. I had to get a Montel slide in after putting up a Modo one. But, yeah, it's across Europe. Now, constraints. I was crying in my water when I saw the constraints stuff earlier, but I'm going to give it a different angle. So that's pretty much the same slide as Lewis showed. This is taken from the NISO's weekly OTF. I don't know if many of you sit in it. It's a really good view of what the grid has been doing over the week. I commend it to anybody. They won't answer your questions, but it's still a good session. Bizarrely, one of the later connectors we had put in, which came from Norway, comes in between B6A and B7, which seems to me a bit mad, given they've already got a problem with power coming from north to south. They could have just extended it 50 miles. They've come 700 miles already, or whatever it was. But, anyway. But this is a really important graph, because of something that's going on. Now, you've heard about Bussuos costs. You've heard about balancing costs. And one of the things that's really adding to those costs at the moment is something called repetitive retrading. How many people know what that is? You're in the business. I mean, people who aren't. Yeah. So, repetitive retrading. There's a consultation out about it at the moment by Desnes, is it, I think? Or is it Offgem? Offgem. Yeah, good. Anyway, there's a consultation about it. Because what's going on here is that batteries that are sited behind those constraints have come up with this sort of cunning plan of when they know the constraint is running, and you can see that these capacities are quite constrained, when they know the constraint's running, they trade ahead in the day-ahead market, setting themselves to export, knowing that they will not be able to export because of the constraint. And the next day, in the balancing mechanism, bid themselves off. And they keep on doing that. And this costs quite a lot of money. So, repetitive retrading is specifically targeted at batteries to try and stop them doing that. And there are about four options out there. There's a consultancy out there, and you can comment on it. We've got our own thoughts on what's the way out of it. But it's basically exploiting a market inefficiency to, ultimately, all of our costs through Bessuos. And that's just a little stat showing how each of the three of the key boundaries of sort of the capacity has moved around through the summer. It doesn't help. And, obviously, in the summer, they tend to be doing a bit of maintenance on those. But it was really interesting seeing those bootstrap diagrams, really, because that's going to get around a lot of this problem and might solve some of that repetitive retrading problem on its own. And that was my back-of-the-fag-packet calculation. I can't attribute it because it was a back-of-the-fag-packet calculation. So, it might be right, might be wrong. It's about that. Skip rates. Now, skip rates are, again, probably some of you aren't totally familiar with skip rates, but it's really important to remember that in all the markets that we have and the auctions we have, which are market mechanisms, a balancing mechanism isn't a market mechanism. It is a balancing mechanism, which means that it's not in merit, by which I mean that the grid won't automatically take the cheapest offers and the most expensive bids when it's trying to trade out, because mostly it's trying to solve balancing its system. So, it means that sometimes if you've got an asset which can do balancing services, for example, we've got one in Gloucester, the grid might ignore it totally because it's just not helping with it. So, if you're sort of thinking about how you're going to be running your battery and you leave yourself to a balancing mechanism, which is really, in some ways, your last chance saloon to get some revenue for that half hour, and you get skipped, you're going to be a bit sick. And skipping for batteries is particularly heavy. Blue being the batteries on that and all the other colours being other technologies. Battery owners are understandably a bit miffed by this. There's a piece of legislation in place at the moment called GC166, which is gradually being deployed, in which the grid gets better view of batteries' state of energy and therefore can have more confidence in the battery's ability to provide services for the balancing mechanism. And that's been largely why they have been being skipped up to now because the grid can't really see how much power and energy the battery's got. GC166 should improve that. It's gradually being rolled out, so it'll be really interesting to see how these stats change. But as you can see, 39% is a big percentage, and it moves around by month, pretty much between 35% and 45%. Right. Then, just when you might be getting confident about the skip rates, how long are you going to be running for in a settlement period? The answer is not very long. And as you can see, the average instruction is about five minutes. You might get two instructions in a half hour, but even then that's ten minutes. So when you're thinking about optimising between the balancing mechanism of wholesale trading, wholesale trading, I know the amount of power I'm going to be delivering and energy I'm going to be delivering. I know it's going to be half an hour. Balancing mechanism, I haven't got a clue. And as I say, it's the last chance saloon. So instinctively, my revenue certainty is greater, even though some of the pricing and balancing mechanism we saw from that earlier graph looks more appealing. It's the lack of certainty that it does for you. Now, the reason I showed that graph earlier, and I said it's not as simple as it looks, because it showed the intraday trading as a single line. And it clearly isn't a single line because the agile trader has a lot of opportunities in one half hour to get in and out a few times. And remember, you're holding a battery because it's a volatility asset. So you're looking to exploit every bit of volatility up to the last minute. And as you can see, you might have sold this, it's the evening peaks, you've probably sold it day ahead. You come in, you've got a chance to get out of that. It will have started probably around 150, because that's when trading started, there was nothing left, but that would have been close to the day ahead price. You'll have had a chance to get out at about 80, around half past seven in the evening. And lo and behold, late doors, you get a chance to sell again and actually dispatch. So this is what you've got to keep in mind. If you've got a battery, is trade it as much as you can. This is, just to give some context, we took a day in the life of Farnham, again using fabulous Montel graphics, which are, as you know, always super detailed. So once you can pick the bones out of it, you won't be able to, just in this short presentation, but I'll give you a quick summary. This battery mostly did, it charged itself using dynamic containment high, those green blocks at the top, the very top, don't worry about the bottom half, during the day. In the morning peak, it did some selling in the wholesale market. Then all day, I think that orange one is dynamic regulation high, it charged. And then in the evening, sold a bit, did a bit of dynamic containment low. And in the evening, charged again using dynamic moderation. Now, I've got a couple of observations about this. One, as I said, you've got to spend quite a long time doing frequency responses to get any power in or out of your battery. And you've got to make a good estimation of what you think your state of charge is going to be at the end. So, it does come with an amount of lack of certainty. But oftentimes, frequency response can give you cheap power or expensively good delivery. But it lacks certainty. But interestingly, this battery, I had another look at it yesterday when the prices were getting really high because I thought, oh, it's got, it must be trading now. It's not going to be doing frequency response. But weirdly, it was doing something called slow reserve most of the day. And I was baffled. And I looked at the operator whose name should, I don't want to shame anybody in public. They might make electric cars. But anyway, but, and they were, most of their fleet actually was doing slow reserve. I think on a day when prices were that good, doing slow reserve. So, I suppose the lesson is, before I finish off, because I can see you lurking, the lesson is, is have a good look at your operator and ask them about their strategy and how they think about the market. Because otherwise, you might find yourself doing stuff that just doesn't make any sense to me anyway. Right. That was it. Thank you very much. Thank you.